gamedge VS tivi

To view the whole list of issues, take a look at the following link: https://github.com/mars885/gamedge/issues

You can take a look at my project called Gamedge. I try to keep it as up-to-date as possible and currently In the middle of converting it to Jetpack Compose.

7. Gamedge

PR migrating from Moshi to Kotlinx.Serialization.

If anyone want to take a look, here's a link to the GitHub.

Look no more: https://github.com/chrisbanes/tivi

It's definitely the most verbose part of our app and one of the weakest parts of the Jetpack library. We used https://github.com/chrisbanes/tivi for some inspiration on how to organize your navigation structures.

I have considered a few variants around this route. One easy way out would be just to launch a ComponentActivity with a PreferenceFragment like what Chris Banes does in TiVi ... but if you go with PreferenceFragment(Compat) it seems you are kinda stuck with SharedPreferences. If one attempts to supply DataStore wrapped in PreferenceDataStore, it seems types get messed up (TextPreference will store as a String, even if the InputType is number, etc).

Thanks I think what's confusing is that UI events are recommended to be modeled as Flows, which I guess is why Tivi uses Flows for even the simplest actions, but it does a lot of very fancy things. Most repository interactions seem to have an entire class extending an Interactor (e.g.) which catches and emits errors, progress and handles timeouts and does all that crazy queuing stuff with UI "messages" Flow

I would read and study nowinandroid and tivi apps.

Check out the approach from tivi https://github.com/chrisbanes/tivi/blob/main/app/src/main/java/app/tivi/AppNavigation.kt as an idea.

They probably want you to do a module split like in https://github.com/chrisbanes/tivi or https://github.com/igorwojda/android-showcase, but what they probably don't realize is that modularization is a trade-off, namely greatly increased build configuration complexity for... well, having modules.